the collaboration copyrighted material imperative · another cutthroat motive was suggested by the...

P1: OTA/XYZ P2: ABCc01 JWBT082-Phelps February 24, 2009 5:15 Printer Name: Yet to Come

Chapter 1

The CollaborationImperative

O n Sunday, May 25, 2003, I was playing golf near myhome in New Canaan, Connecticut, when I receivedan unexpected phone call.

“Hi, Marshall, this is Bill Gates,“ said the caller. “I knowthat Brad [Smith, Microsoft’s general counsel] spoke with youyesterday about the offer. But I just wanted to reinforce our hopethat you’ll come to Microsoft and help us with this really bigchallenge that we’re facing.”

Bill and Brad had already outlined the nature of that chal-lenge when I met with both of them nine days earlier at the com-pany’s Redmond, Washington, headquarters: a limited patentportfolio that failed to protect Microsoft’s huge R&D invest-ment or provide it with the new business opportunities createdby today’s fast-changing technology environment. In short, theysaid, Microsoft needed a first-class patenting program and an

1

COPYRIG

HTED M

ATERIAL


B U R N I N G T H E S H I P S

intellectual property (IP) strategy that could facilitate the closecollaboration with other firms that Microsoft needed to succeedin this new landscape of business competition.

“I know you’re enjoying your retirement now,“ Bill went on.“But I really believe you’re the person with the right backgroundto handle this job.”

I told him that I’d have to talk to my wife first, but that theopportunity did indeed sound exciting.

“That would be great,” Bill replied. “We’re all familiar withthe great work you did at IBM, and I’m really looking forwardto working with you.”

It appeared that Bill had read some of the press reports onmy work at IBM, which noted how (to quote one report) Ihad “put IP on the corporate map and made senior manage-ment and Wall Street sit up and take notice of IP as a revenuegenerator.” During my 28-year career at IBM, I had led thetransformation of the company’s patent licensing program intoan almost $2 billion per year profit machine—more profit justfrom IP licensing, it should be noted, than the total earningsof all but the top 40 largest companies in America at the time.As one of the first senior executives in corporate America tosee profit and competitive advantage where others had seenonly legal documents sitting in the filing cabinets of corpo-rate law departments, I had helped to kick-start a revolutionin the way that companies manage their intellectual propertyportfolios.

Three days after the phone call, I met with Bill and Bradagain. And over the course of several more days of discussion,we reached agreement on the scope of my responsibilities andthe company’s commitment to this effort. On June 5, 2003,Microsoft announced that I would become the company’scorporate vice president of intellectual property.

2


The Collaboration Imperative

The announcement had a rather electrifying effect. As oneIP trade journal put it, “When the world’s richest man hires thearchitect of the world’s most lucrative intellectual property pro-gram, the [business world] takes notice.” And when the world’srichest man happens to also be regarded in some circles as theworld’s biggest monopolist, it’s no wonder that his hiring of ahigh-profile patent “warrior” might have caused some alarm.

“The fact that Microsoft hired Marshall Phelps tells youeverything you need to know about their intentions,” insistedanalyst Russell Parr in an interview with MSNBC at the time.According to Parr, I had been brought to Redmond to recreatethe massive $2 billion-a-year IP royalty stream I had built forIBM in the 1990s.

Another cutthroat motive was suggested by the technologyand business magazine ZDNet: “Microsoft is very keen to [usepatents to] rein in Open Source,” it argued, referring to the freesoftware movement. “Marshall Phelps will do that.”

Not for the first time, of course, the pundits were wrong.The idea that a significant industry force such as the open sourcemovement could ever be hemmed in by me or anyone else waspatently absurd. And in point of fact, a key objective of the li-censing program we planned to launch was to build a cooperativebridge to the open source world in order to meet customer de-mands for greater interoperability between Windows and Linuxsoftware.

As for trying to recreate IBM’s $2 billion-a-year IP royaltystream, suffice it to say that for a company that generates abillion dollars in free cash flow every month, a mindless focuson maximizing licensing income made no business sense at all.In all my talks with Bill and Brad, both before and after I washired, we never once discussed the idea of building an IBM-stylerevenue juggernaut from patent licensing. Instead, we merely

3



hoped that licensing might generate sufficient revenue to coversome or all of the costs of maintaining the patent portfolio.

What none of these nervous analysts seemed to realize wasthat my hiring had come at a moment of profound change forthe company. For years, Microsoft had been on the defensive,beset on all sides by antitrust suits and costly litigation, andviewed by many in the technology industry as a monopolistand market bully. At the same time, the dynamics of technologydevelopment and the software business had begun to change rad-ically, requiring Microsoft to adapt to the emerging era of “openinnovation,” in which collaboration between firms, rather thango-it-alone market conquest, would be the keystone of success.

Underlying Microsoft’s desire for change was its recognitionthat technology development had become too widely dispersedand heterogeneous, the pace of innovation too rapid, and thecompetition for markets and customers too multifaceted and de-manding for any one firm to go it alone anymore. Indeed, it wasbecoming increasingly difficult for even the largest companies tohold all the pieces of even their own product technology in theirown hands anymore. This was true even for Microsoft, whichinvested billions of dollars a year in research and development.In this new, decentralized technology environment, therefore,companies like Microsoft would simply have to collaborate ifthey wanted to succeed.

Bill, Brad, and I were hardly the only ones who had cometo this conclusion. Indeed, in an era in which inventions werebuilding upon each other with such rapidity as to quickly renderobsolete even the most far-sighted company’s product develop-ment strategy, it was becoming clear to a growing number ofbusiness leaders that the only way to stay above the rushingwaters of creative destruction was to stand on the firmament ofalliances with other firms. Whereas some 80 percent of major

4



innovations during the 1970s had come from inside a singlecompany’s own R&D labs, by the dawn of the twenty-firstcentury, studies now showed, more than two-thirds of ma-jor new innovations involved some sort of interorganizationalcollaboration—either between private firms, or between firmsand federal laboratories or research universities. Recognizingthat fact, 7 out of 10 senior executives surveyed by The Economistwould conclude that their best strategy for accelerating innova-tion was to increase collaboration with other firms.

As my colleague Masanobu Katoh, then-corporate vice pres-ident for intellectual property at the Japanese giant Fujitsu,recently noted: “We are a $45 billion company. We are intoconsumer products, computers, consulting, and services, evenmanufacturing. We do it all. But alas, doing it all is no longerenough. We can no longer succeed unless we collaborate withother companies.”

This new “collaboration imperative,” as I called it, was re-shaping business and redefining the sources of competitive ad-vantage. And more to the point, it was rewriting the rules thatbusinesses have always followed for how they leverage and deployintellectual property.

First and foremost, it meant that intellectual property couldno longer be viewed solely as a negative right—meaning, the rightto either prevent someone from using your technology and com-peting in your market, or to tax them in the form of licensingfees for the right to do so. From now on, IP’s greatest valuewould lie not so much in being a weapon against competitors,but rather in serving as a bridge to collaboration with other firmsthat would enable companies to acquire the technologies andcompetencies they needed to compete successfully.

Indeed, intellectual property was becoming the sine qua nonof open innovation itself. It provided the legal scaffolding upon

5



which firms could share their most innovative research and part-ner together to create new products and services. Without IPrights, firms would resist sharing their ideas for fear that otherswould misappropriate their innovations. But with such rights,firms could share their innovations with others, secure in theknowledge that all were fully protected in deploying them tomutual advantage. Just as good fences made good neighbors,strong IP rights would make for strong and successful collabo-rations. And in the case of intellectual property, this so-called“fence” would turn out to be more of a bridge than a barrierbetween firms.

We Need Relationships

This was the vision that informed Microsoft’s new direction. Be-cause collaboration now appeared to be the key to Microsoft’sfuture success, the company’s greatest need was to start buildingrelationships with other firms—large firms, small firms, opensource firms, venture capitalists, software developers, even inde-pendent inventors. In short, Microsoft needed relationships withanyone and everyone it could find in order to remain at the cen-ter of technology innovation and at the forefront of new marketsand business opportunities. And intellectual property was quitesimply the best available vehicle for constructing those collabo-rative relationships. As the legal embodiment of any company’smost precious resource—its innovation—Microsoft’s IP definedthe arena of cooperation and established clear rights and obliga-tions for both sides in any joint endeavor.

Commenting on my hiring, the industry trade journal eWeeknoted that “Gates agreed with Phelps that Microsoft was at aninflexion point in its history” and needed to cooperate morewith other firms. Newsweek magazine, meanwhile, reported that

6



I took the job “only after Gates promised [me] he wantedto change the way Microsoft interfaced with the technologyworld.”

It’s important to recognize that my hiring was only thelatest in a whole series of moves on the part of Microsoft’ssenior leaders to repair the company’s relations with the in-dustry and government. A year earlier Bill Gates and CEOSteve Ballmer had appointed Brad Smith as the firm’s new gen-eral counsel, based, in part, upon Smith’s insistence that “itwas time to make peace” and establish more mutually bene-ficial relations with other companies. And Smith had alreadymoved swiftly to translate that vision of peace and collabora-tion into reality by resolving dozens of legal cases against thecompany, including Microsoft’s 2002 antitrust settlement withstates’ attorneys general, its data privacy agreements with theFederal Trade Commission and the European authorities, and,later on, antitrust and intellectual property disputes with AOLTime Warner, Sun Microsystems, RealNetworks, IBM, andNovell.

As Ballmer told BusinessWeek: “The company has made it apriority to do all we can to end these legal issues and to do soin a way that increases collaboration with other companies.”

The company’s efforts in this regard were profoundly impor-tant in setting the stage internally for Microsoft’s new intellectualproperty and technology initiatives, even before my arrival. AsBrad Smith would later tell a Harvard University forum com-memorating the tenth anniversary of the U.S. Justice Depart-ment’s antitrust suit against Microsoft, “Part of our maturationprocess really required that we start to see ourselves the wayother people were seeing us. It was not necessarily an easy thingto do. It seldom is when you go through that kind of process.But it spurred us to take a more principled approach. It enabled

7



us to do business in a different way. And I think it has served usand the industry well.”

I must confess that I was humbled by the prospect of helpingsuch a powerful high-tech leader to develop a new vision for thecollaborative use of intellectual property in the age of open inno-vation. Microsoft is a company, after all, with an enormous foot-print in the global information technology industry—a high-tech Bigfoot if ever there was one.

Microsoft itself employs 95,000 people worldwide. Butconsider that of the 35 million people employed worldwide asof 2007 in the $1.2-trillion-a-year information technology (IT)sector, an astonishing 42 percent—or 14.7 million people—areemployed either directly by Microsoft or by firms that makeproducts that run on Microsoft software, provide support forMicrosoft software, or otherwise participate in the global Micro-soft “ecosystem.” This includes 181,200 Microsoft-based jobsin Malaysia (44 percent of Malaysia’s IT ecosystem), 71,607Microsoft-related jobs in Ireland (47 percent of Ireland’s ITworkforce), and 4.2 million Microsoft-related jobs in the UnitedStates (42 percent of our domestic IT workforce).

What’s more, the 640,000 vendors in this global Microsoftecosystem earned more than $425 billion in revenue in 2007,invested close to $100 billion in local national economies, andtheir employees paid more than $514 billion ($203 billion in theUnited States alone) in taxes that year.

In short, Microsoft is a vital engine of the global economy.And if it were going to embark on a new initiative to share itstechnology and intellectual property with any and all interestedfirms—including competitors—and launch technical and marketcollaborations with both the largest multinational enterprises aswell as the smallest startups and entrepreneurs, this could nothelp but have profound effects on the shape and direction of the

8



whole technology industry. What’s more, I was acutely awarethat our efforts to adapt to the new “collaboration imperative”would inevitably offer lessons—both positive and negative—notonly for IP managers but for senior executives throughout theinformation technology sector as well.

But to understand where Microsoft needed to go, it wasnecessary to first appreciate where it had been—and why, in theview of many inside and outside the company, the company haddeveloped a “fortress mentality” towards its technology that itnow needed to overcome.

A Lesson in Patent Holdups

One of the most insightful historians of Microsoft’s developmentis Nathan Myhrvold, Microsoft’s chief technology officer (CTO)from 1986 to 1999, and later the founder of the invention andintellectual property licensing firm Intellectual Ventures. A bril-liant mathematician, amateur archeologist, and world-class pho-tographer, Myhrvold has an uncensored ebullience that comespartly from his personality, and partly from being a billionaire.

“When I first came to Microsoft in 1986, patents meantnothing to the company,” recalls Myhrvold. “I think they hadtwo patent applications, but most people didn’t even know whatthe word ‘patent’ meant. For one thing, software wasn’t widelyunderstood to be patentable back then. And within the softwareindustry as a whole, there wasn’t any focus on it. We were stilljust trying to legitimize the notion that software itself was a realbusiness apart from hardware.”

(Actually, the legal basis for patenting software was estab-lished in the 1981 Supreme Court decision Diamond v. Diehr,which affirmed that software and computer-implemented in-ventions could be patented. Software patenting didn’t become

9



widespread until the Federal Circuit Court’s Allapat decision in1995, which led the U.S. Patent and Trademark (PTO) office toissue Examination Guidelines for Computer-Related Inventionsa year later.)

“But over time,” Myhrvold adds, “I began to realize theimportance of doing more advanced research and development,and of securing patents for our discoveries so we could con-trol our own technological destiny. In 1991, we finally foundedMicrosoft Research, which is now the largest private researchorganization in the world—bigger even than Pfizer’s, Merck’s,or even IBM’s research arm. And we started filing more patents.”

But Myhrvold soon realized just how deep Microsoft’s patentdeficit actually was. “As Microsoft got bigger,” he explains, “allsorts of companies started coming around to see us. They’dclaim that we were infringing their patents, and demand that wetake a license. And I was like, ‘Oh my God, they can do this?They can just demand money from us?’ A lot of people wereshocked by that, I can tell you. And when our lawyers lookedaround and asked what sort of patents we could assert backagainst these companies—in a sort of ‘mutual assured destruc-tion’ show-down that would enable us to cross-license withouthaving to fork over a lot of money—the answer was, ‘We don’thave crap.’ So every time one of these companies came by toassert their patents against us, it would cost us money. Sometimes50 or 100 million dollars. And that’s a lot of zeroes to give awayjust because someone else has patents and you don’t.”

At times the patent gold digging took on comic overtones.“There was this one guy named Efraim Arazi, who started apioneering computer graphics company called Electronics forImaging, or EFI for short,” recalls Myhrvold. “This also hap-pened to be his nickname—Efi—so obviously this was a guywith no small ego. But that’s true of a lot of geniuses, and Efi

10



was a genius. He was a founder or cofounder of several compa-nies, and an outstanding engineer. But his most brilliant idea wasto license this seminal patent from MIT that was the result ofeight years and $4 million worth of research into color imagingand printing technology. A total of four Ph.D. and 12 mastersdegrees were awarded to researchers who contributed to thatproject. And Efi licensed the patent that resulted from all thiswork for only $25,000, plus royalties.”

Myhrvold smiles at the memory. “So Efi takes this patentand starts making the rounds. He shows up at Kodak about oneweek after that famous Polaroid judgment [in which Kodak wasordered to pay Polaroid $925 million for patent infringement,the largest patent award in U.S. history at the time]. And in thatinimitable style of his, he says—and here you have to imaginehis Israeli accent—‘So, Kodak, you’re not doing so well, eh?Well, look at this patent I have here.’ And Kodak says, ‘Oh God,not again!’ and immediately writes him a check for $10 million.Then he goes to Adobe with this same patent and makes theman offer they can’t refuse, either. So that‘s another $10 million.And finally he comes to Microsoft.”

Myhrvold laughs at the memory of Arazi’s audacity. “So Efiand I sit down in a room with our patent attorney, and he startstelling us the story of the $20 million he‘s already gotten forhis little $25,000 patent and how the money is funding his newcompany. ‘Hey, this is better than getting money from venturecapitalists,’ he gushes. ‘I don’t even have to give up any equity!’

“And then all of a sudden,” adds Myhrvold, “Efi turns toour patent attorney and says, ‘So, Mister Patent Attorney, didyou know Windows is infringing my patent?’

“Well, our patent attorney, who was this very buttoned-down guy, suddenly doesn’t look so well,” Myhrvold explains,“so he leaves the room to go to the bathroom. As soon as he’s

11



out of the room, Efi starts in on him. ‘Oh my God, he’s going togo home, he’s going to go slap his wife and yell at her for buyinghim that stupid tie!’ And I can’t help myself. I’m laughing myhead off. I mean, Efi really was a character.”

But after sharing a good laugh, Myhrvold says he got serious.“I sort of cleared my throat and told Efi very firmly that unlikeKodak and Adobe, we didn‘t have color management. So hecould go pound sand as far as I was concerned.”

Efi’s response? “‘Okay,’ Efi said, ‘no problem.’ He was like,‘Hey, it was worth a try!’”

But perhaps the most painful lesson in Microsoft‘s earlypatent education was delivered by a small company named StacElectronics. Explains Myhrvold: “Stac had a disk compressionprogram for our old DOS operating system called Stacker. Andin 1991, I think it was, we started negotiating with Stac to seeif we could license their compression program and integrate itinto the new 6.0 version of DOS we were about to release.Well, we couldn’t reach a deal—I don‘t remember what theproblem was—so instead we simply decided to design aroundStac’s patent. Sure, it cost us more money to design our owncompression tool, which we called DoubleSpace, than if we hadjust licensed it from Stac. But since we couldn’t get the license,we had no choice.”

Insists Myhrvold: “We really did design around their patent.We came up with a whole new way to achieve disk compressionthat was different from the way they did it. Stacker used X, Y, andZ methods to achieve compression, whereas our compressionprogram used P, Q, and R methods to get a similar result. Whichis exactly what we’re supposed to do under patent law. Webelieved it was perfectly legal. We felt very confident that wedidn’t infringe their patent, and that we could prove it. So,we figured, game over.”

12



Except that the game was not over. “When we releasedDOS 6.0 with that DoubleSpace feature, they sued us forinfringing their patent anyway,” Myhrvold recalls. “Not onlythat, but while we were working to design around their patent,Stac had gone out and bought this old data compression patentowned by Ferranti International Signal, a British aerospacecontractor, which had developed it for a satellite system. It wasoriginally filed as a hardware patent—this was before softwarepatents were even allowed—but the claims were written broadlyenough that a court could interpret it to apply to software aswell. Stac paid a quarter of a million dollars to buy that littlepatent. And they bought it specifically to use against us.”

Myhrvold sighs ruefully. “I was totally shocked. I rememberasking our attorneys, ‘They can do this? They can just go buya patent and then use it retroactively against us?’ And they said‘Yep.’ And I remember, I just went, ‘Aw, hell . . .’”

He shrugs. “So Stac won a judgment against us for $120million. The jury said we infringed. We appealed, but latersettled with them for something like a $40 million invest-ment in their company and an additional $43 million in roy-alties. All because of a patent they bought for a quarter milliondollars.”

It would later turn out, of course, that Stac’s patent pur-chase was merely a harbinger of the sort of wholesale purchasingof patent rights for assertion against deep-pocketed firms thatwe see everywhere today. This can be an especially egregiouspractice, bordering on extortion, when committed by firms thatdon‘t produce any products—the infamous “patent trolls.”

“We were driving over patent landmines every day,”Myhrvold remembers. “It was terrible. And it wasn’t just us.The whole software industry paid no attention to patents inthose days. You know, back then we were all just hell-bent on

13



shipping product next week. So the Stac suit really drove it hometo me that Microsoft was in real danger on the patent front. Itwas a major lesson to me.”

And for Bill Gates as well. Indeed, it was precisely in responseto the ever-growing number of patent holdups by firms such asStac that Microsoft after 1993 began requiring all its partnersand OEMs to agree to a non-assertion of patents (NAP) clausein their deals. The NAP clause would later prove to be a majorbone of contention between Microsoft and the industry—and itsabandonment ten years later by Microsoft would likewise proveto be the key factor in enabling the company to forge a new andhealthier relationship with the industry—but it’s important tounderstand the defensive origins of Microsoft’s “fortress mental-ity” practices.

In the words of Horacio Gutierrez, who in 2006 would suc-ceed me as vice president of intellectual property and licensingat Microsoft, “You have to remember what it was like for thiscompany in those days. We were this odd group of people com-peting against the biggest forces in the high-tech world to createthis new software industry. It was not necessarily obvious at thattime that software was a viable business apart from hardware.And when you’re fighting for your survival like that, you havea tendency to be protective—especially if you’re a start-up ina new emerging business sector. It was only later, as we ma-tured, that we began to realize—and our customers began to tellus—that we didn’t need to be so defensive anymore and that,with our growth, we had to assume a larger responsibility forthe health of the whole industry.”

In fact, it is precisely Microsoft’s ability to change that im-presses Gutierrez most about the company: “The only con-stant that you find in the history of this company is this almost

14



relentless willingness to re-examine ourselves,” he insists. “That,and a willingness to reinvent ourselves and shed the things thatare preventing us from moving forward.”

In any event, these patent holdups eventually got Myhrvoldthinking about Microsoft‘s future as well as his own. “I realizedthat patents were becoming incredibly valuable in today’s world,and they were going to be vital to any company that wanted tocontrol its own technological destiny.” So convinced was he ofthe importance and value of patents—not only to business, but toAmerica’s innovative capacity and competitiveness as well—thathe eventually resigned from Microsoft in 1999 and started hisown firm, Intellectual Ventures (IV), to invest in invention andthe development of a liquid market for patents.

This is how I came to know Myhrvold—by becoming anearly partner in Intellectual Ventures as well as IV’s negotia-tor with Microsoft on its 2003 investment in the IV inventionfund. One week I sat on IV’s side of the table, discussing invest-ment terms with Microsoft negotiators John Weresh, a leadingIP attorney, and Kenneth Lustig, one of the company’s top deal-makers and strategists in the corporate development group. Thenext week, after agreeing to take the Microsoft job, I was on theMicrosoft side of the table, as Weresh and Lustig’s boss.

In fact, the very day I started at Microsoft, Bill and Brad di-rected Weresh and Lustig to go ahead and get the IV investmentclosed and obtain the best possible terms for Microsoft, whichthey did. The IV investment was very important to Microsoft,given the increasing numbers of patent suits it faced as well asthe company’s paramount interest in seeking opportunities tocollaborate with others. According to Lustig, “We began to re-alize during the course of our negotiations with IV that therewere some very positive opportunities that could be realized by

15



embracing IP as a business in a thorough and cross-functionalway. Critical to realizing these opportunities would be the build-ing of wide-ranging IP relationships in the industry, with bothpartners and competitors alike, to reduce litigation conflicts andwork toward a common set of interests.”

In any event, shifting from one side of the table to theother left me with an odd feeling—rather like some sort ofhemispherical brain malfunction. And it made me realize yetagain how small the world really is.

Small, perhaps, but no less challenging for the intimacyof my connections. My job was to reform and invigorateMicrosoft’s IP department, which in my view had been func-tioning like a football team composed only of defensive linemen,with no one knowing how to throw a forward pass. To be sure,there were people in other corporate groups within Microsoftwho were trying to play offense and build partnerships, but theseefforts were not broad enough. And no one was yet system-atically looking for ways to employ the company’s intellectualproperty as the glue for cementing sturdy partnerships withother firms.

This would have to change. I knew that we had todevelop an intellectual property strategy that could help thecompany achieve its most critical business objectives. Thesewere to transform Microsoft’s relations with the industry so thatwe could better participate in the new and more distributedenvironment for technology innovation, pursue joint develop-ment efforts with other companies, gain access to technologywe needed, and pursue joint marketing and sales opportunitieswherever possible. But how exactly to do all of this, I wasn‘tquite sure.

Which, come to think of it, is exactly the same spot I wasin just a decade earlier.

16



It’s Deja vu All over Again

In hindsight, it’s now clear that my rather eclectic series of jobs atIBM was leading me inexorably towards the use of intellectualproperty as a financial and strategic asset in IBM‘s recoveryfrom its near-death experience. But at the time, I could see noclear pattern in my dissatisfied wanderings through the IBMbureaucracy. All I knew was that I wanted to do something thatcombined legal and business activity. But what that was, I hadno idea.

I started out in the legal department in 1970 and held avariety of positions there. But I was never totally sold on stayingwithin legal. For one thing, IBM’s legal department had a repu-tation as an organizational black hole—a “roach motel” that letyou check in but not check out. So I was thrilled when, aftersome major antitrust litigation that I worked on came to a close,my bosses decided to send me to Stanford business school. Thecompany paid all the bills—my apartment, telephone, nurseryschool for the kids, my salary, the whole bit. It was a great gig.

When I graduated from business school in 1980, I went towork for the then–chairman of the company, Frank Cary. Now,in IBM’s universe, becoming an assistant to the chairman orpresident was considered the best training with an eye towardsthe future. You studied at the feet of the master, so to speak,working on whatever projects he happened to be focused on. Itwas considered a springboard to a higher management career.

Over the next 13 months, I worked very closely with bothCary and IBM‘s general counsel—the former U.S. attorneygeneral and undersecretary of state under President Johnson,Nicholas deB. Katzenbach—on two key issues: how to settlethe Justice Department antitrust suit against IBM, and what thefuture organization of IBM should look like, beyond the issues

17



that could arise if we lost the antitrust suit. We already had twoplans—a Red Plan and a Blue Plan—for splitting IBM into twodifferent companies if we lost the suit. Our work focused onhow we could rationalize the company if we won or otherwisesettled the antitrust suit and remained a single enterprise. At thetime, IBM was plagued with a duplication of structures and ef-forts. These parallel universes were developed to protect IBM ifthe company was ever broken up. As a result, we had laboratoriesand plants whose main competitors were other IBM laboratoriesand plants.

It was a great learning experience for me. We visited variousIBM laboratories and factories all over the world, trying to figureout whether we had the right structures and the right divisionsin place. I didn’t realize it at the time, but this experience wouldprove to be enormously helpful to me later on.

After that, I held various legal positions over the nextfew years—including general counsel for the personal com-puter business at IBM and managing attorney for the lawdepartment—but again, I had the education and the trainingand most of all the desire to do something more, something inthe business arena. Finally, in 1984, I was asked to help lead thecompany’s Asia business as vice president of IBM’s Asia/PacificGroup.

My new assignment to Japan was actually part of a majorcorporate restructuring in which IBM hived off a piece of itscorporate headquarters and sent it to Asia. The reason for thiswas really quite simple, although younger executives may find ithard to understand nowadays. Recall that this happened beforethe Internet and even before the widespread use of faxes, at atime when it might take all night for a teletype machine to crankout a few pages of a document. As a result, we found it was verydifficult to run a business that stretched from India in the west,

18



to Japan in the east, to Korea in the north, and then all theway south to New Zealand—all from corporate headquartersin Armonk, New York. We were constantly putting people onairplanes and flying them back to headquarters to discuss whatsort of bids we were going to make for various pieces of theAsia mainframe business. And we discovered that, as often as not,by the time the bid details were worked out and the negotiatorsflew back to the Far East, we had lost the deal.

So we needed a new way to run the Asia business. And intypical IBM fashion, we did it juggernaut style. We moved 500families to Japan, something no other American corporation hadever done. One of the Japanese newspapers called it “McArthur’sSecond Invasion.”

“You’re Stealing Our Software!”

At about the same time as my posting to Tokyo, IBM began tosuspect that the Japanese manufacturers who built clones of ourmainframe computers—Hitachi, Mitsubishi, and Fujitsu beingthe three biggest ones—were illegally copying our mainframesoftware.

According to the consent decree IBM signed with the U.S.Justice Department in 1956, we were required to license ourintellectual property to everyone in the world on reasonable andnon-discriminatory terms. But after 1969, when we unbundledour software from the hardware and sold it separately, it was ourview that Japanese mainframe manufacturers needed to createtheir own software for these machines. Yet despite being rela-tive newcomers to mainframe software, their software somehowmanaged to accurately match the capabilities of our own.

So we commissioned some of our top programming and se-curity experts to do some detailed forensic work on the Japanese

19



software. That’s when our suspicions quickly gave way to cer-tainty. Apparently, one of the programmers in Poughkeepsie,New York, who had designed the IBM software, named oneof the loops in the program after his girlfriend, whose name,if I recall, was Judy. And our forensics revealed that the samename—“Judy”—appeared in the source code of the Japanesesoftware. Now Judy is not a very common name in Japan, sothis turned out to be the silver bullet that proved the Japanesehad infringed our software intellectual property.

We brought the matter up with Fujitsu, Mitsubishi, andHitachi, and the latter two firms settled rather quickly. Butnegotiations with Fujitsu were another matter entirely. On theIBM side we had Ralph Pfieffer, the head of IBM World Trade,and John Opal, who would later serve as IBM’s CEO. ViceChairman Michio Naruto led the Fujitsu team. We would meetin a suite at one of the major hotels—usually the Okura, becauseit was centrally located—and the negotiations quickly becamevery difficult, and very personal, as well. These were not civildiscussions at all. I recall one incident in which the two sidesliterally screamed at each other from across the room.

“Fujitsu is acting immorally,” Ralph Pfeiffer charged.“You’re stealing our software!”

Vice Chairman Naruto bolted out of his seat: “You Amer-icans are just trying to keep Japan down! You don’t like it thatyou are no longer the masters in our relationship.”

It’s important to understand some of the geopolitical dy-namics that were contributing to the animosity in that room, forit was not simply these two companies, Fujitsu and IBM, whowere responsible for the hard feelings. The 1980s were a verydifficult period in U.S.-Japan relations. Japanese industrial andfinancial power was clearly ascendant, whereas the United Stateshad lost much of its competitive edge, not only in consumer

20



electronics but in a number of manufacturing sectors as well. Toput it bluntly, Japan was beating the crap out of America in avariety of competitive arenas. And we were scared.

In any event, after years of negotiations and arbitration,Fujitsu finally settled under terms similar to those agreed toby Hitachi and Mitsubishi. A substantial sum of money was in-volved that was nonetheless smaller than the amount they wouldhave had to pay to develop their own software. And today, iron-ically, not only does IBM cooperate with Fujitsu in a variety ofendeavors, but I am also proud to call Naruto-san, now retiredas vice chairman of Fujitsu, a very good friend. Whenever I amback in Tokyo, we get together and laugh about old times.

I learned many things during that trying time in Japan.I developed a deeper appreciation for the diversity of humanexperience and the importance of trying to see the worldthrough other-than-American eyes. But even more important, Ireceived my very first lesson in the enormous value and impor-tance of intellectual property to corporate, and indeed national,success—especially as it related to software technology.

I carried this lesson back home with me when in 1987I was appointed IBM’s director of governmental programs inWashington, D.C. This was a public policy job in which mytask was to promote government and industry cooperation toenhance technology innovation and the competitiveness of U.S.industry around the world. I had seen how Japanese industry andgovernment cooperated to strengthen that country‘s competitiveposture in the world. And we looked for ways that we might beable to achieve some of the same result, albeit in a way that bettersuited our own nation’s more laissez faire economic system.

During this time, we helped form the Council on Compet-itiveness, which played a crucial role in spotlighting issues of na-tional competitiveness and innovation for both government and

21



industry policymakers. Along with my counterparts at Hewlett-Packard, Sun, and other firms, I helped organize the first “CEOForum” in 1989 under the auspices of the Computer SystemsPolicy Project, which brought the top people from the dozenlargest computer makers (the “hardware guys”) together for thefirst time. These people were not used to working together, orindeed even talking to each other, on matters of joint concern,so we didn’t know how the event would turn out. In fact, I wasquoted in BusinessWeek as saying, “We weren’t sure these guyscould have dinner together without getting into a food fight.”But in the end, it was successful in not only establishing tiesbetween America’s high-tech CEOs that have persisted to thisday, but also in lobbying Washington to regulate Japanese tradeand dumping practices.

Again, you have to remember that this was a time of greatconcern in the United States about the state of American in-dustry. The number of new inventions being patented by U.S.technology companies was in rapid decline. Even IBM, whichhad previously been a leading U.S. patentee, was in danger offalling off the top-ten patenting chart completely, as were othergreat American companies such as General Electric. In contrast,Japanese companies such as Fujitsu, Mitsubishi, and Hitachi hadsoared to the top of the patenting and innovation chart.

The bottom line was that we once thought we had theinformation technology world to ourselves, only to discover thatour king-of-the-mountain position on the high-tech landscapewas being cratered by the Japanese. And looming on the horizonwere the Taiwanese and the Koreans.

In retrospect, of course, it’s now clear that we failed toappreciate both the vitality and resilience of U.S. innovationas well as the thin reed that the Japanese model of competi-tiveness, with its tight industry-government cooperation, was

22



really built upon. We also didn’t realize the extent to which thatJapanese industrial juggernaut was fueled by a huge real estate andfinancial bubble that would ultimately collapse by the end ofthe 1990s.

In any event, after four years in Washington, I was ready tomove on. It wasn’t because I thought my policy work aroundcompetitiveness wasn’t important. Rather, it’s that I’m basicallyapolitical—I’m neither a Democrat nor a Republican—and itseems to me if you’re going to be in a political position like that,you ought to know if you’re a boy or a girl, so to speak.

So when IBM’s then-chairman, John Akers, came toWashington in late 1991 to lobby government officials on somematter, he and I went to dinner and he offered me one of threecorporate vice-presidencies. One choice was to stay where I was,only with additional responsibility as a corporate vice president.Another was vice president of finance for the PC business. Andthe third was vice president of commercial and industry rela-tions, or C&IR. For reasons far too Byzantine to explain here,C&IR and not the law department was responsible for IBM’sintellectual property portfolio, among other duties. I decidedto take the C&IR job, where I felt I could put my growinginterest in technology innovation and intellectual property tosome use.

IBM on the Ropes

It turned out that I took over the intellectual property portfolioat a crucial time for IBM, because the company was literallyon its deathbed. Over the next two years, IBM sustained $15.4billion in losses—it lost $8.1 billion in 1992 alone, the largestcorporate loss up to that point in U.S. business history—andsales of its old standby mainframes, from which 90 percent of

23



the firm’s profits were derived, had dropped by half. By the timeJohn Akers retired and the board hired Lou Gerstner to take overas CEO in April of 1993, the company had fewer than 100 daysof cash remaining in the bank.

The reasons for Big Blue’s decline were varied, but takentogether they pointed to a bureaucratic lack of innovation andan arrogant refusal to adjust to changing market conditions.The company had completely missed the minicomputer trend,was a laggard in the PC business, was losing mainframe marketshare hand over fist, and was saddled with a costly overheadthat included all sorts of duplicative programs and a policy oflifetime employment that prevented IBM from weeding outunder-performing managers. Instead of innovating, all we weredoing was hemorrhaging red ink and propping up outmodedbusiness models.

So the first thing I did as vice president of C&IR was tochange its name to Intellectual Property and Licensing (IP&L)and shut down its industry relations and other programs. I laidoff about half of my organization—probably close to a hundredpeople. Of course that was only a drop in the bucket comparedto the retrenchment going on throughout the company at thattime. Indeed, the layoffs and restructuring would eventually trimthe IBM workforce to half its former size. We went from 407,000employees in 1986 to barely 200,000 a decade later.

It’s one thing to get rid of a deadwood operation, of course,and quite another to actually do something that positively en-hances the company’s business prospects. So I began to lookclosely at the economics of the company’s R&D and its rela-tionship to our patenting efforts. We were spending more andmore every year on R&D, yet filing fewer and fewer patents toprotect that R&D investment. So I decided that IBM was goingto start patenting aggressively again. We set a goal of becoming

24



number one on America’s patenting chart. We made it by 1993,just a year and a half after I took over the IP function, and IBMhas remained number one ever since.

Filing thousands of patents, of course, is not cheap. So Ilooked for ways to transform the economics of our effort so thatit would become a profit center rather than a cost center for thecompany. I realized that licensing our intellectual property wouldbring in royalties from companies that used our technology, butI wasn’t sure how much. One thing I was sure of, though,was that the only way to get IBM’s divisions to go along withlicensing technology to other firms, including competitors, wasto share the proceeds with them. So I decided that every dollarof royalties that we brought in from licensing should go rightback to the divisions.

Now, that was a very radical idea in American business atthe time—not just sharing your technology with others, whichwas considered heresy in many quarters, but using the proceedsfrom such licensing to not only fund the patenting and licensingoperations themselves but also to enhance the bottom line of thecompany‘s divisions. But I was able to get away with it for onesimple reason—the old CEO, Akers, was on his way out, butthe new CEO, Gerstner, was not yet fully in charge. I was ableto do it because no one was there to stop me.

To make this licensing effort successful, however, I knewI had to find a partner among the divisions to help develop aproof of concept. So I went to see Mike Attardo, the chief ofthe microelectronics division, which was then going througha very rough time with the transition from bipolar to CMOS(complimentary metal oxide semiconductor) chip technology.

“Mike,” I told him, “you’re laying off all these bipolar engi-neers. Why not give me 10 of them—along with some scanning,tunneling microscopes so they can analyze the chips being made

25



by our competitors? And I’ll bet you I can make 25 millionbucks for your division.”

Now, this was a pretty attractive offer for a guy like Attardo,given that his division, like most at IBM at the time, was onthe ropes financially. I mean, here’s some guy offering him $25million, for which he doesn’t have to do a damn thing exceptlend some engineers and some microscopes? He said fine, no skinoff his behind. And that first year we made him $48 million.

How did we do it? We used these scanning, tunneling mi-croscopes, which can examine objects at the molecular level,to look at the chips made by our competitors. And lo and be-hold, we discovered that in a number of cases the grooves thatwere etched in competitors’ chips—they’re called “vias”—wereexact, and I mean exact, replicas of the grooves used inIBM chips.

Now, it takes a very precise chemical process to etch chipsin exactly the way that IBM had etched them. There‘s only oneway to do it, in fact, and we had patented it. So we used thesemicroscopes to photograph the vias in other companies’ chips,and then we’d go meet with those companies and suggest thatmaybe they needed to take a license to our technology.

There we’d be, all of us sitting around a conferencetable—lawyers and engineers from both of our companies—andit was the same process in every meeting we held. We’d ask theother side to please explain to us how the vias in their chipscould so perfectly match our own, down to the molecular level.Because, so far as we knew, there was only one way on earth todo it, and we had patented that method. And there would be anawkward silence from the other side.

Did anyone ever admit to infringing our patents? No, ofcourse not. But a couple of weeks later, the other companywould get back to us and agree that perhaps a license between

26



our two companies would be a good idea after all. Eventually,they all took licenses.

Now, it should be noted that the licensing deals we signedwere not confiscatory. Besides ensuring that other firms paid forour technology when they used it, we usually tried to licensethem additional technology or know-how so that they felt thatthey were getting a good deal. In fact, this licensing effort ofours was one of the first examples in American industry ofcompanies licensing and sharing technology with other firmsfor mutual benefit.

What’s more, while I ran the IP organization for IBM, Inever sued anybody (and I take pride in the fact that I rarelysued when I later ran the IP organization at Microsoft). Thisis no small feat in an age when patent litigation has become aburden on business. Of course, sometimes you have to litigatewhen the other party has stolen your hard-earned innovationand refuses to resolve the problem. And even a non-litigiouscompany like Microsoft has had to sue one or two companies inthe past couple of years, but only as a last resort. But in general,I believe the world would be a lot better off if more companieswould treat their intellectual property primarily as a businessand financial asset and not a litigation club for beating damageawards out of rivals.

In any event, after our success with the licensing program inthe microelectronics division, word started getting around to thevarious divisions that we could make real money for them. Andgiven that we had already spent the money to do the underlyingR&D, the revenues from licensing were 98 percent pure profit.So suddenly other divisions wanted to work with us to licensetheir technology as well. This became a virtuous circle as therevenues raised became part of the following year‘s divisionalbusiness plans.

27



But while support for our program was growing, there wasa lot of pushback as well. For one thing, the divisions tendedto see the technology they developed as belonging to them, notto the whole company. And they viewed the whole notion oftechnology sharing as heresy. After all, throughout the entirehistory of American business, the goal had always been to keepyour technology from your competitors. So licensing was viewedas a radical new idea.

There was also, as you would expect, a good deal of op-position from some of the companies we targeted for licensing.Those that were IBM mainframe customers would often respondby calling an IBM sales manager and complaining about the waythey were being treated. Then that IBM sales manager wouldcall me and scream about how I was screwing up their salesrelationships with these firms. Sometimes both of them wouldcomplain directly to the CEO’s office, which would then callme and demand an explanation.

Convincing the Big Boss

Most of these complaints I was able to beat back. But when Igot a call from Lou Gerstner himself on perhaps the second orthird day after he took command at IBM, I worried that perhapsthe jig was finally up.

“What the hell are you doing?” he demanded.“Er, um, what do you mean?” I replied.“Are you licensing our technology to competitors?”“Yes, I am,” I told him.“So then I ask you again,” he sighed, obviously frustrated.

“What the hell do you think you’re doing? Why aren’t you usingour patents to stop our competitors in their tracks?”

28



There it was again—the idea that intellectual property wassimply a weapon for blocking other companies rather than afinancial asset and a vehicle for collaborating with other firms.

“Lou,” I said, “let me come to your office and prove to youwhy it’s a good idea.”

“Fine,” he said, “I’m eager to hear it.”I hung up the phone and turned to the two colleagues who

were sitting in the office with me. “Now what?” I asked them.“How do we do this?”

So we started batting ideas around for how to make the casefor licensing to Lou. I have to admit that I was more than alittle nervous. After all, Lou Gerstner was an outsider—he hadcome from RJR Nabisco and American Express—and he hadabsolutely zero loyalty to anyone in the corporate bureaucracywho didn’t pull his or her own weight. We had no doubt thathe was going to fire every single corporate officer in IBM ifthat’s what it took to get the company out of its doldrums (and,in fact, he eventually did fire almost every corporate officer inheadquarters). So we couldn’t bluff this one. We couldn’t justuse some pretty overlays with pie charts to convince him. Wehad to come up with something tangible, something real andundeniable, to demonstrate that our licensing program was vitalto IBM’s survival and recovery.

And then one of the guys—I forget exactly who—said,“Why not crack open a laptop and show him everything thatwe’ve licensed?” So that’s what we did. We pried off the top of anIBM laptop, and the keyboard case as well, and then we glued abunch of little map flags—toothpicks with red flags on them—toevery single technological component in the computer that wehad licensed from other companies. After a couple of hours,we had 150 of these little flags sticking out from almost everysquare inch of that laptop. And the only reason we stopped there

29



is that we simply ran out of room. And bear in mind, this wasIBM’s own computer architecture—and yet IBM still neededthe intellectual property of other firms to build it.

Later that afternoon, we brought the laptop to Lou’s office.My memory of events is a bit sketchy after 16 years, but the gistof what happened is as follows:

“Take a look at that,” I told him. “We can’t even build ourown computers without other people’s technology. The wholetechnology world is interdependent now, and there’s no goingback. There’s no way we’re ever going to get out of the licensinggame.”

Lou was silent, still staring at that laptop with all those littlered flags.

“So here’s our idea, Lou,” I continued. “If we have to licenseanyway just to stay in business, why not make some moneyfrom it?”

With less than a hundred days cash left in the company, ofcourse, that got his attention.

He looked up at me. “How much money?” he asked.I paused, took a deep breath and looked up at the ceiling.

“A billion dollars,” I said. “We’ll make you a billion dollars bythe end of this decade.”

He looked at me very hard then. I could tell he was debatingwhether to ask me how I came up with that ridiculous billion-dollar number, but he decided to let it pass. Which was a luckything, because I had totally grabbed that number out of thin air. Isuspect he probably guessed as much, but since no one else in thecompany was promising to make him a billion dollars, I supposehe decided what the hell, let this Phelps character go for it.

“What do you need?” he asked.“Just leave me alone,” I replied, “and I’ll make you that

billion dollars. But I’ve got to have the authority to do what I

30



want with the IP. I can’t have the business managers telling meit belongs to them and their divisions. And I can’t have the salesmanagers complaining every time I try to get one of their bigcustomers to take a license from us.”

Lou nodded quietly, then said “Okay.”And sure enough, he later sent an e-mail to everyone in the

company—it was probably his very first e-mail at IBM, if notthe first e-mail he had ever sent in his life—informing them thatthe intellectual property of the various divisions belongs to theIBM corporation, not to the divisions. And that sole authorityfor its use belongs to the IP and Licensing group.

“Intellectual property assets such as patents belong to IBM,not to the individual units,” Gerstner wrote. “It is one of our keycorporate assets and must be protected. Negotiations concern-ing intellectual property with companies outside IBM are theresponsibility of the Intellectual Property and Licensing staff.”

In the end, we reached a billion dollars in cumulative li-censing revenues—not in seven years as I had promised Lou,but in three years. In fact, we earned a phenomenal $1.9 billionin the year 2000 alone, which was when I retired from IBM.And 98 percent of that was pure profit—all except for the $36million it cost me to run the department. Bottom line, our in-tellectual property-licensing program contributed 25 percent ofIBM’s total profit for that year.

IP Must Serve the Business

Many analysts and reporters have written about the successwe achieved in building such a lucrative intellectual property-licensing program at IBM. But all too often, they have missedthe real lesson from our success. They have talked about the IPlicensing effort as a business in its own right, and this had the

31



unfortunate effect of contributing to a mistaken belief on thepart of some companies that they, too, could replicate IBM’sbillion-dollar IP revenue stream. Those who tried to do so gen-erally failed.

The point being that intellectual property should always servethe business, not be the business. When I started the licensingprogram at IBM, Big Blue was on the ropes financially, with onlya hundred days worth of cash on hand. Its most critical businessneed was revenue, and that’s what we designed our IP strategyto generate. But when I went to work for Microsoft a decadelater, Microsoft didn’t need money—it had billions of dollarsof cash in the bank. Instead, Microsoft needed to transform itsrelations with the rest of the industry and build collaborativerelationships with other firms. So that became the focus of ournew IP strategy.

Which sounded good in theory. But almost immediatelyafter I accepted the IP job at Microsoft in June of 2003, thattheory ran up against the harsh reality of the NAP (non-assertion of patents) clause—the same NAP clause that Gatesand Myhrvold had designed to help save the company 10 yearsearlier. How we dealt with the new realities of NAP would bethe true test of our new collaboration strategy.

32

the collaboration copyrighted material imperative · another cutthroat motive was suggested by the...

Documents